Résultats de la recherche - decoding (algorithmicsics OR (algorithm OR (Algorithmssssssics OR (algorithm OR algorithm))))

Guess-Verify-Refine: Data-Aware Top-K for Sparse-Attention Decoding on Blackwell via Temporal Correlation par Cheng, Long, Zhao, Ritchie, Liu, Timmy, Li, Mindy, Qiao, Xianjie, Duan, Kefeng, Chen, Yu-Jung, Chen, Xiaoming, Rouhani, Bita Darvish, Yang, June

Table des matières: “…Sparse-attention decoders rely on exact Top-K selection to choose the most important key-value…”
Accéder au texte intégral

VeriCache: Turning Lossy KV Cache into Lossless LLM Inference par Yao, Jiayi, Shen, Samuel, Du, Kuntai, Feng, Shaoting, Seo, Dongjoo, Zhang, Rui, Huang, Yuyang, Liu, Yuhan, Lu, Shan, Jiang, Junchen

Table des matières: “… the high decoding throughput of a range of KV cache compression algorithms. VeriCache uses the compressed…”
Accéder au texte intégral

Implementation and Optimization of HQC Decoding on NPU-Integrated Devices par Chau, Vu Minh, Kiet, Nguyen Ngoc, Minh, Pham Quang, Ngoc, Mai Xuan, Anh, Nguyen Duc, Ta, Hoang

Table des matières: “…-based key-encapsulation mechanism, providing algorithmic diversity alongside lattice-based post-quantum…”
Accéder au texte intégral

FBCNet: A Multi-view Convolutional Neural Network for Brain-Computer Interface par Mane, Ravikiran, Chew, Effie, Chua, Karen, Ang, Kai Keng, Robinson, Neethu, Vinod, A. P., Lee, Seong-Whan, Guan, Cuntai

Table des matières: “… by Motor Imagery (MI) decoding algorithms for electroencephalogram (EEG) based Brain-Computer Interface…”
Accéder au texte intégral

TOFU: Towards Obfuscated Federated Updates by Encoding Weight Updates into Gradients from Proxy Data par Garg, Isha, Nagaraj, Manish, Roy, Kaushik

Table des matières: “… and privacy, we propose TOFU, a novel algorithm which generates proxy data that encodes the weight updates…”
Accéder au texte intégral

Best Arm Identification Based Beam Acquisition in Stationary and Abruptly Changing Environments par Ghatak, Gourab

Table des matières: “… propose a novel algorithm called concurrent beam exploration, CBE, in which multiple beams are grouped…”
Accéder au texte intégral

Masked Transformer for Electrocardiogram Classification par Zhou, Ya, Diao, Xiaolin, Huo, Yanni, Liu, Yang, Fan, Xiaohan, Zhao, Wei

Table des matières: “…. With the advent of advanced algorithms, various deep learning models have been adopted for ECG tasks. However…”
Accéder au texte intégral

Q-Segment: Segmenting Images In-Sensor for Vessel-Based Medical Diagnosis par Bonazzi, Pietro, Li, Yawei, Bian, Sizhen, Magno, Michele

Table des matières: “… present "Q-Segment", a quantized real-time segmentation algorithm, and conduct a comprehensive evaluation…”
Accéder au texte intégral

JAMDEC: Unsupervised Authorship Obfuscation using Constrained Decoding over Small Language Models par Fisher, Jillian, Lu, Ximing, Jung, Jaehun, Jiang, Liwei, Harchaoui, Zaid, Choi, Yejin

Table des matières: “…, a user-controlled, inference-time algorithm for authorship obfuscation that can be in principle…”
Accéder au texte intégral

Demystifying the Physics of Deep Reinforcement Learning-Based Autonomous Vehicle Decision-Making par Wan, Hanxi, Li, Pei, Kusari, Arpan

Table des matières: “… in AVs. Therefore, in this research work, we try to decode the knowledge learnt by the attention-based…”
Accéder au texte intégral

LLMServingSim: A HW/SW Co-Simulation Infrastructure for LLM Inference Serving at Scale par Cho, Jaehong, Kim, Minsu, Choi, Hyunmin, Heo, Guseul, Park, Jongse

Table des matières: “… (LLM) inference serving systems. These efforts not only include innovations in the algorithm…”
Accéder au texte intégral

A Deep Generative Learning Approach for Two-stage Adaptive Robust Optimization par Brenner, Aron, Khorramfar, Rahman, Sun, Jennifer, Amin, Saurabh

Table des matières: “…. In this work, we introduce AGRO, a solution algorithm that performs adversarial generation for two-stage…”
Accéder au texte intégral

MixNet: Joining Force of Classical and Modern Approaches Toward the Comprehensive Pipeline in Motor Imagery EEG Classification par Autthasan, Phairot, Chaisaen, Rattanaphon, Phan, Huy, De Vos, Maarten, Wilaiprasitporn, Theerawit

Table des matières: “…-computer interface (BCI) systems, enhancing the decoding of electroencephalography (EEG) signals. However…”
Accéder au texte intégral

Hardware Acceleration of Kolmogorov-Arnold Network (KAN) for Lightweight Edge Inference par Huang, Wei-Hsing, Jia, Jianwei, Kong, Yuyao, Waqar, Faaiq, Wen, Tai-Hao, Chang, Meng-Fan, Yu, Shimeng

Table des matières: “…, MUXs, decoders, etc.). For the first time, this paper employs an algorithm-hardware co-design…”
Accéder au texte intégral

A Generative Framework for Predictive Modeling of Multiple Chronic Conditions Using Graph Variational Autoencoder and Bandit-Optimized Graph Neural Network par Rico, Julian Carvajal, Alaeddini, Adel, Faruqui, Syed Hasib Akhter, Fisher-Hoch, Susan P, Mccormick, Joseph B

Table des matières: “… decoder, are then processed by a GNN using a novel Laplacian regularization technique to refine the graph…”
Accéder au texte intégral

Deep Transfer Learning-based Detection for Flash Memory Channels par Mei, Zhen, Cai, Kui, Shi, Long, Li, Jun, Chen, Li, Immink, Kees A. Schouhamer

Table des matières: “…) algorithm for flash memory channel detection. It can effectively reduce the training data size from $10^6…”
Accéder au texte intégral

AADNet: An End-to-End Deep Learning Model for Auditory Attention Decoding par Nguyen, Nhan Duc Thanh, Phan, Huy, Geirnaert, Simon, Mikkelsen, Kaare, Kidmose, Preben

Table des matières: “…Auditory attention decoding (AAD) is the process of identifying the attended speech in a multi…”
Accéder au texte intégral

Towards Flexible 3D Perception: Object-Centric Occupancy Completion Augments 3D Object Detection par Zheng, Chaoda, Wang, Feng, Wang, Naiyan, Cui, Shuguang, Li, Zhen

Table des matières: “…-centric occupancy perception from both data and algorithm perspectives. On the data side, we construct…”
Accéder au texte intégral

Optimum Power Allocation for Low Rank Wi-Fi Channels: A Comparison with Deep RL Framework par Mohsin, Muhammad Ahmed, Bhattacharya, Sagnik, Rajabalifardi, Kamyar, Pote, Rohan, Cioffi, John M.

Table des matières: “… algorithm for multi-carrier NOMA, named minPMAC, and an associated time-sharing algorithm that adaptively…”
Accéder au texte intégral

Upside Down Reinforcement Learning with Policy Generators par Di Ventura, Jacopo, Ashley, Dylan R., Herrmann, Vincent, Faccio, Francesco, Schmidhuber, Jürgen

Table des matières: “… this using Hypernetworks - a variant of Fast Weight Programmers, which learn to decode input commands…”
Accéder au texte intégral